Jump to content
UltravioletPhotography

Sony A7III conversion to multispectral


JMC

Recommended Posts

A few months back I got a Sony A7III as I was finding myself in need of a mirrorless camera for some imaging work. As the A7III has a BSI sensor I thought it would be an interesting one for conversion to UV imaging, given it should have increased short wavelength sensitivity.

 

I've done some preliminary testing on it here;

 

https://jmcscientificconsulting.com/project-mirrorless-part-1-the-beginning/

 

And here (after the conversion);

 

https://jmcscientificconsulting.com/imaging/project-mirrorless-part-2-the-camera-conversion/

 

One thing I did do is as well as having the ICF stack removed, I had the sensor coverglass replaced with fused silica, as the Sony one blocks loads of UV starting at 400nm. I've been seeing this with Canon and Nikon too.

 

Initial tests are promising, at 310nm the A7III is about 1 and 1/3 stop faster than my converted 5DSR, while at 365nm they are about the same. A small but useful improvement when trying to see shorter wavelengths. The Bayer filter is robbing the full potential from the BSI sensor for imaging in this region though.

 

Plenty more testing to go, and I hope to get some nice comparison images in the new year.

 

Oh, and Merry Christmas everyone :)

Link to comment

Thank you for these tests.

I think the biggest problem for us going into the future is that coverglass from the leading sensor manufacturers is blocking UV.

That is a big concern I have for future cameras. With all the added advantage of BSI sensor will be lost due to the coverglass.

However, for regular photography it may mean getting rid of the thick UV/IR stack. So lens design can be simpler, like in the film days with no glass on top of the sensor. Old lenses will be better and old cameras will be better for UV photography.

Link to comment
Colin, in my limited experience with Canons, my experience has been that the microlenses are beneficial even in the UV (the focussing they provide is better than the amount of UV they absorb). Whether this is true for all sensors I cannot say.
Link to comment
Do you have a feeling about if the UV-cut of this camera's sensor's cover-glass is more aggressive than something like in your older Canon DSLR?
Link to comment

Ulf, I have the data somewhere. I'll try and get it together on one graph and post it on here.

 

EDIT: I've dug out a couple of spectra for sensor coverglasses. One for the Canon 5DSR and the other for the Sony A7III. I have a coverglass for a Nikon d800 somewhere, but haven't run it yet. Both severely limit UV transmission especially at short wavelengths.

 

post-148-0-05357200-1577699253.jpg

Link to comment
  • 2 weeks later...

JMC, when you're saying you had the sensor coverglass replaced, do you mean the thin layer of glass that's directly on top of the sensor?

I did't know this could be removed.

And did you replace the ICF stack after removing it?

I'm going to modify an A7 soon and some people say you have to replace the ICF stack after removing it because otherwise you have to recalibrate focusing by adjusting the spring loaded adjustment screws on the sensor board. However, I don't see the logic in this since focusing is done on the main sensor, not a separate focusing sensor like in DSLR cameras.

Link to comment

Hi Nisei, yes the thin layer of 'glass' on top of the sensor was removed in addition to the stack. The thin layer of 'glass' was then replaced with fused silica. Nothing else was added. I say 'glass' like that as I'm fairly sure it isn't glass, the cutoff looks too gradual. MaxMax did the conversion for me, and you're right this coverglass isn't normally removed, and it is a very delicate job.

 

I've not had an issue with focusing and MaxMax did not recalibrate that as far as I know.

Link to comment

That's good to hear.

I'm confident in that I can remove the filter stack (did it with a Nex-3N as well) but replacing it with clear glass without having a single speck of dust getting caught between the sensor and glass sounds like a near impossible task without a clean room.

Link to comment

Thanks for the graphs Jonathan. I'm tempted to have a A7III converted too, but not with your extreme coverglass mod.

 

The need for readjustment in some way after removing the ICF-stack is due to that the optical path-length is changed without the stack.

In some cases you might lose the ability to focus at infinity as the path is getting longer.

Link to comment

The need for readjustment in some way after removing the ICF-stack is due to that the optical path-length is changed without the stack.

In some cases you might lose the ability to focus at infinity as the path is getting longer.

I've never had any infinity focus issues with my Nex-3N but I was mostly using that with a lens adapter that has a filter inside (so behind the lens). Perhaps this compensates enough for the missing stack.

I hope just removing the stack on the A7 won't give me issues when I use front filters though.

I've specifically chosen an A7 model from the first series because A: there's a step by step tutorial on how to remove the filter stack on Lifepixel and B: because the II and III series have IBIS, the sensor unit is much more delicate and I don't dare to pry off the stack that's securely held in place with double sided tape.

But perhaps there are people here who've been brave enough to do the modification themselves?

Link to comment

I've never had any infinity focus issues with my Nex-3N but I was mostly using that with a lens adapter that has a filter inside (so behind the lens). Perhaps this compensates enough for the missing stack.

I hope just removing the stack on the A7 won't give me issues when I use front filters though.

I've specifically chosen an A7 model from the first series because A: there's a step by step tutorial on how to remove the filter stack on Lifepixel and B: because the II and III series have IBIS, the sensor unit is much more delicate and I don't dare to pry off the stack that's securely held in place with double sided tape.

But perhaps there are people here who've been brave enough to do the modification themselves?

 

Rear-mounted filters do compensate well if they have an optical path-length similar to the original internal filter stack.

With front mounted filters you might get problems with infinity focus after the conversion.

Link to comment
  • 1 month later...

I got my converted A7III back from having its electronic view finder fixed, so I can start some more tests now.

 

One interesting result today. I repeated part of the ISO test I did on my monochrome d850, which is discussed here (https://www.ultravioletphotography.com/content/index.php/topic/3730-multispectral-monochrome-nikon-d850-testing-and-shakedown/page__view__findpost__p__33300).

 

With the A7III, I left the body cap on, and did a series of exposure 1s in length, from ISO200 to ISO102400. None of them showed any fogging. There was of course more noise across the whole image at the higher ISO settings, however no issue with a fog at one side like with the d850.

 

Must try some high ISO photos in the UV.....

Link to comment

Could your fog issue be a light leak? Did you wrap the body and lens with foil and test for one?

The d850 tests were done in the dark Andy so the ones I shared were not light leaks. I do however think it leaks some light as well, perhaps near or by the LCD screen on top, as I've seen additional issues when using it outside. More tests needed there though to verify.

Link to comment
Be aware of a possible leak area around the "prong" on the aperture ring, for Nikkor AI/AIS lenses. It's not the prong as such that can cause a light leak, but the section of the aperture ring that is cut away on these lenses.
Link to comment

The d850 tests were done in the dark Andy so the ones I shared were not light leaks. I do however think it leaks some light as well, perhaps near or by the LCD screen on top, as I've seen additional issues when using it outside. More tests needed there though to verify.

I know the A7II series had this problem because Sony was using a UV LED inside the camera as part of the autofocus system. On converted cameras this was causing problems with longer shutter speeds.

Perhaps Nikon used similar electronics in the D850?

Link to comment
  • 3 years later...

Hi JMC, I noticed in your Sony A&SIII UV mod you had them replace the IR cutoff/OLPF with a fused silica window? Did Max Max supply that window or did you have to provide it for them?

Link to comment
1 hour ago, DKoch said:

Hi JMC, I noticed in your Sony A&SIII UV mod you had them replace the IR cutoff/OLPF with a fused silica window? Did Max Max supply that window or did you have to provide it for them?

It was supplied by Maxmax. Technically though, it was the sensor coverglass that was replaced with fused silica, not the IR cutoff filter pack. However I would not recommend you use the fused silica windows for sensor coverglasses unless you really need to go below 300nm, and a Schott WG280 coverglass will certainly let more UV through than the one that is on the camera as standard. The fused silica windows tend to be dirtier or more prone to surface defects than normal Schott WG280 and I have had to have one of mine with fused silica replaced (I went back to standard WG280). The WG280 is what MaxMax normally uses when he does his monochrome conversions.

Link to comment

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...